Scale your Teradata infrastructure to deliver massive improvements to capacity and
performance by disaggregating compute and storage with MinIO's high performance, AWS S3
compatible, distributed object store.
Teradata's Native Object Store is a new offering to expand the amount of data available to their flagship Vantage analytics platform. The NOS initiative has multiple components including disaggregation, pay-as-you-go pricing and expansion to public and private object stores. MinIO is a launch partner for Teradata’s NOS initiative and brings a differentiated suite of features to Teradata’s performance-driven customer base.
Teradata customers now have a complete view of their business with the ability to see data on their S3 compatible object stores. Object storage is the de facto class in the cloud, and Teradata Vantage can now read from this massive store of semi-structured and unstructured data to drive business insight and value.
MinIO's NOS implementation scales seamlessly from a single rack of six Petabytes to hundreds of Petabytes spanning multiple data centers and geographies in a single namespace. This provides a massive data store for Teradata Vantage to analyze.
The economics and scalability of object storage are well understood - particularly given MinIO’s cloud native architecture. By combining simplicity and scale, Teradata Vantage customers will deploy less capital while simultaneously spending less time managing infrastructure.
MinIO's exceptional throughput characteristics make it the fastest NOS on the market. Teradata customers can design their private cloud object stores using standard servers from leading providers. Choose HDD or NVMe to create the optimal mix of cost and performance while enjoying read speeds in excess of 55 GB/s on a small 8 node NVMe cluster.
By creating tables against an S3 resource, one can simply upload objects, which dynamically adds them to the table. These can be queried directly via Teradata.
By using Teradata NOS and MinIO together, one can join table against S3 compatible sources allowing you to query existing data alongside data in the object store.
Because data can be loaded directly as objects, there is no need for complex or time consuming ETL jobs.
By enabling per-bucket access policies through MinIO, one can allow upload and sharing privileges across multiple organizations.